This section describes the sampling frame, the sampling process and calculation of the sample size.
3.4.1 Sampling Frame
The sampling frame was developed by listing the names of all pre-school teachers working in the two selected Sub-Counties of Kisii County.
Table 3.1: Distribution of Pre-school Teachers in the Sampling Frame
Gender
Pre-school Teachers by Sponsor ECD Centre
Pre-school Teachers by Location of ECD Centre
Public Private Total Urban Rural Total
Men 12 08 20 10 10 20
Women 194 150 344 217 127 344
3.4.2 Sampling Techniques
Orodho (2004) points out that a sample is selected in such a way that one is assured that certain sub-groups in the population will be represented in the sample in proportion to their numbers in the population itself to avoid bias in the sample drawn. Taking into account important factors such as available resources and the desired groups for the study, purposeful sampling was used to select the area of study. That is how Kisii Town, Nyaribari Chache and Mosocho Sub-Counties were selected to represent the nine sub-counties of Kisii County. As for the study respondents, a simple probability random sampling was used. Every name in the sample frame was written on a piece of paper and then segregated according to gender. The pieces of paper were then wrapped and shuffled. From each of the two groups (men and women), the names of the respondents were randomly selected. The process was then repeated, this time according to the location where the pre-school teacher worked, namely; rural and urban locations.
3.4.3 Sample Size
In survey design studies, Mugenda and Mugenda (1999), and Bryman (2008) recommended that the minimum sample size should be at least 10% for a large sample and 20% for a small one. On the strength of that recommendation, the researcher drew a sample size equivalent to 20% of the number of pre-school teachers in the study area (refer Table 3.1). Ultimately, the calculated sample size was 73 (Table 3.2). However, 72 respondents returned their questionnaires.
Table 3.2: Proportional Sample Size from the Sample Frame
3.5 Research Instruments
The researcher developed and used two instruments to collect raw data, namely; a structured questionnaire and an observation checklist.
3.5.1 Structured Questionnaire
Orodho (2004) defines a structured questionnaire as that having closed or enclosed questions with appropriate options from which a respondent selects the answer that best describes his or her situation. Bearing this in mind, the researcher constructed a semi-structured questionnaire based on the study objectives. Part one of the questionnaire comprised the respondent‟s demographic characteristics (numbers 1 to 3) while in part two (numbers 4 to 22) related to teachers‟ job satisfaction.
3.5.2 Observation Checklist
The observation checklist was constructed because of the need, as recommended by Mugenda and Mugenda (1999), to use multiple methods in data collection in which case the merits of one instrument may offset the demerits of the other. The checklist was used to collect observed data based on Gender
Pre-school Teachers by Sponsor
Public Private Total
Men 2 2 4
Women 39 30 69
the pre-school teacher‟s job environment. It measured variables such as the size of classrooms, availability of teaching materials and play equipment.
3.6 Pilot Study
The questionnaire was pre-tested using respondents from two randomly selected pre-schools within the two sub-counties of Kisii County. The said respondents were excluded from the study. The pilot study enabled the researcher to familiarize herself with the anticipated challenges arising from logistics and data collecting instruments. For instance, the researcher was able to moderate on the average time it took the respondent to fill the questionnaire and to determine how to respond to common questions raised.
3.6.1 Validity of the Study Instruments
Mugenda and Mugenda (1999) define validity as the accuracy and meaningfulness of inferences, which are based on research results. According to Orodho (2004) validity, is essentially concerned with establishing whether the questionnaire content is measuring what it is supposed to measure. Validity, in this case is a non-statistical measure of the degree to which several measures of the concept, accurately measure the concept. Content validity involves the examination of test content to determine whether it covers a representative sample of the behaviour domain to be tested; representative validity involves turning an abstract theoretical construct to a specific test practical (Anastasi & Urbina, 1997). To ensure content and representation validity of the research instruments, the researcher sought expert review by three judges who were experienced in social research.
3.6.2 Reliability of the Study Instruments
Reliability is a measure of the degree to which a research instrument yields consistent results or data after repeated trials (Saunders, Lewis & Thornhill, 2009). To check for consistency in content, the researcher performed the split- half test of reliability (Ndani and Kimani, 2010). To do this, the questionnaire was administered to 8 pre-school teachers, randomly selected from a frame of ECD centres in the area of study. Scores from even numbers were correlated with scores from odd numbers using Spearman‟s Correlation Coefficient, a non-parametric statistical measure whose values range from minus one (-1) to plus one (+1). The value of the test co-efficient (r) was 0.6 (r = 0.6), indicating a positive reliability of the questionnaire.